Monday, December 29, 2008

The twelve puzzles of Xmas

White to play. Can he win?


Tom Chivers said...

Mm. Not if he goes directly for the a-pawn with his king...

Jack Rudd said...

1.Ke6 Kg3 2.Kd5 Kf4 3.e5! does it, yesno? That way, after 3...Bxe5 4.Kc6, black can't play 4...Ke5, and ends up a crucial tempo behind.

Anonymous said...

Perhaps, but wouldn't be easier just to steam forward e6-e7 then Ke6 bagging that vital tempo straightaway?

ejh said...

Jack - it would, and yet...

ejh said...

(I should note that Jack isn currently occupied trying to checkmate Simon Williams, so don't leave it to the poor chap...)

Anonymous said...

I've now realised the flaw. Black just ignores our threat to take the bishop on e7!

But I have it this time, for sure. We need to combine threats to promote the e-pawn with threats to take the a-pawn, viz:

1. e5 Bb2 (if 1. ... Bxe5 then ignore it! 2. Ke6 -> takes a7) 2. Ke6 Kg3 3. Kd7 Kf4 4. e6 Ba3 5. Kc7 Ke5 6. Kb7 and Black's king is in a right old dilemma. 6. ... Kd6 allows 7. e7 overloading it, and otherwise it can't get back to c7.

ejh said...

(....and even as I write, off go the queens...)

Anonymous said...

Final refinement as we've allowed the drawing ... Bf6 in that line. We should start e4-e5-e6 to force the bishop to a3, and only then run Ke4-d5-c6-b7.

ejh said...

Anonymous...yes, and yet...

Anonymous said...

And yet ... well, the one thing we haven't mentioned is the possible ... Bb8 to free up e5 for the king, viz: 1. e5 Bxe5 2. Ke6 Kg3 3. Kd7 Kf4 4. Kc6? Bb8! 5. Kb7 Ke5 6. Kxb8 Kd6 7. Kxa7 Kc7 draw. But obviously 4. Kc8! enables White to take on b8 en route to a7. Thus the full solution reads

1. e5 Bb2 2. e6 Ba3 3. Ke4 Kg3 4. Kd5 Kf4 5. Kc6 Ke5 7. Kb7 Kd6 8. e7!


1. e5 Bxe5 2. Ke6 Kg3 3. Kd7 Kf4 4. Kc8 Bb8 5. Kxb8 Ke5 6. Kxa7 Kd6 7. Kb7.


I've seen your Tue 30 puzzle before and do remember the rather elegant solution, so will leave that one to your other readers.

ejh said...

That looks pretty good to me. I'll provide a full explanation tomorrow morning. (I might have done it today, but our home PC was rushed into intensive care. It's fine now, and resting.)

ejh said...

The position comes from Gerald Abrahams' Test Your Chess (Pan, 1975) and is the second puzzle in the book, headlines (perhaps with a little unintended irony) Have you ideas?

I'll quote the answer he gives, in full (albeit exchanging his descriptive for algebraic).

This position (from a study by Bondarenko and Liburkin) illustrates levels in chess.

The novice will say: "Yes, e5 wins the bishop. Then the king wins the Black a-pawn and White promotes".

A technician quickly refutes this:

1.e5 Bxd5 2.Kxe5 Kg3 3.Kd6 Kf4 4.Kc7 Ke5 5.Kb7 Kd6 6.Kxa7 Kc7 draws.

Nor does a finesse help. Thus: 6.Kc8 Ke7!

Then is the diagram position a drawn one? Not if you add to your technique an

1.e5 Bxe5 2.Ke4! Kg3 3.Kd5 Kf4 4.Kc6 Ke4 5.Kb7 Kd5 6.Kxa7 wins.

The Black bishop has cost its king a vital move. Nor can the result be avoided. There is a pretty variation:

1.e5 Bc3 2.e6 Bb4 3.Ke5 Kg3 4.Kd5 Kf4 5.Kc6 Ke5 6.Kb7 Kd6 7.e7* and the Black king is deflected from the pursuit.

As we've discovered, this is wrong in a number of respects, although the conclusion, that White wins, and does so by not taking the bishop on e5, remains sound.

However, the whole idea of ...Bb8 has been overlooked, presumably not just by Abrahams but by Bondarenko and Liburkin too, which is odd in a way because it's an attempt to recreate the same idea of boxing in the king as was presented in the first variation. And it's defeated, not by 2.Ke4? (rather than !) but by 2.Ke6 - which move was not considered.

Also see comments nere: one wonders how many old studies, produced before computers, have similarly been found to be flawed.

[* in fact this reads 7.P-Q7, i.e. 7. d7]